From Pixels to Pathways: Omni-Modal Cell Biology with AI
summary
An end-to-end platform called REMI images single cells at 60,000 frames per minute, runs on‑chip AI inference, and physically sorts cells label‑free into buckets while streaming high‑dimensional morphology to the cloud service Axon. At Erasmus Medical Center REMI identifies a rejection‑associated NK cell morphotype nicknamed Mickey Mouse cells, enabling early blood‑based detection of kidney transplant rejection, claiming fewer biopsies and roughly $100M saved in the Netherlands. A unified “GPT for cells” virtual cell model is the stated roadmap.
AI-written summary · about
chapters (6)
- 0:00Introduction to morphomics
- 2:44Morphology as underexploited data
- 5:39REMI platform demo
- 7:48Kidney transplant rejection detection
- 10:39Future of AI-driven biology
- 13:25Conclusion and call to action
related talks
21mFighting Cancer with AI
44mNext-Gen Therapies
22mReimaging Healthcare
39mSustainably Automate Manual Cell-Staining with Innovative Open-Sour...
26mFrom Pixel to Patient: Making Medical AI Work in the Real World
29mWhat do a tree and the human brain have in common-a not so serious introduction to digital pathology
47mBANKSY: A Spatial Omics Algorithm that Unifies Cell Type Clustering and Tissue Domain Segmentation
24mHigh-throughput cellular analytics to identify central nervous system targets for gene therapy